I NIST Calibration of ASTM

E127-Type Uitrasonic
Reference Blocks

J. A. SlotwinskI
G. V. Blessing

U.S. DEPARTMENT OF COMMERCE

Technology Administration
National institute of Standards
and Technology
Ultrasonics Group
Manufacturing Engineering Laboratory
Gaithersburg, MD 20899

—QC
.056
100
NIST
1994
//5430
 NISTIR 5430

NIST Calibration of ASTM

E127-Type Ultrasonic
Reference Blocks

J. A. Slotwinski
G. V. Blessing

U.S. DEPARTMENT OF COMMERCE

Technology Administration
National Institute of Standards
and Technology
Ultrasonics Group
Manufacturing Engineering Laboratory
Gaithersburg, MD 20899

May 1994

U.S. DEPARTMENT OF COMMERCE

Ronald H. Brown, Secretary

TECHNOLOGY ADMINISTRATION
Mary L Good, Under Secretary for Technology

NATIONAL INSTITUTE OF STANDARDS

AND TECHNOLOGY
Arati Prabhakac, Director
 !

"
%'t

toV
vL ' .-

K. >

jg#ig,Sia
(w

•
j-
Irl':-

nnaini:';^
,#'S<'«5«?Ii* > -•' -5'>' .

• :

'.V'. ^
iif
% ,
;

••Is

55s.'

k'*.^
"*
V
*

#
'’^

».
...
''>

'
-

-..,
=*1
<*"
J b
fi*
.
VAiWbaC^
-1^ TvV^V:'
^ LdUb-^ bi
'tM‘!juoe.,.-.,wit^M
;.

J.^
—
IkUkliM •

»
oi^ et

>x #^l
 TABLE OF CONTENTS

Introduction 1

E127-Type Reference Blocks 2

Measurement Facility 2

Instrumentation 3
Transducer 3
Pulser/Receiver 4
PT-4 Tuning Circuit 5
RC Circuit/Digital Voltmeter 6

Calibration Procedures 6
Preliminary Settings 6
Distance-Amplitude (D-A) Sets 7
Area-Amplitude (A-A) Sets 8
Calibration Report Contents 8

NIST Ultrasonic Calibration Databases 8

Calibration Database 8
Check-Standard Database 9

Conclusion 9

Acknowledgments 10

References 11

Appendix I: 13
Procedures for Requesting Ultrasonic Reference Block Calibrations

Appendix II: 15
Data- Acquisition Forms used in NIST Calibrations of ASTM E127-Type Reference Blocks

Appendix III: 19
Sample Distance-Amplitude Calibration Report

Tables 22

Figures 24

1
 I" -r 1- ; >
•

„f

tMfbi^iimT ,JM.^

.' '

,
5
'

Dj^, "H
'
^'V-' ''
’

’
'&'
'}.
,

.:.#^ik,'
" ’

^..At
'

-
;

r ^
n \ - •/. ', ^;!

'^''\‘'^,V' i''V

- - ^
-r-V'J ^
« '

'
'
ats
'
:' '
",
"J
' '

’?
,
,

‘'%;i
fioq&M
Sj^ T. (V*J .
*

l'.i/^.!4
a
:!fr
^
^ V a y%$ha!B.td0o-9^
\‘»'. ^pd'

'•
©1
^
'

'T
7S
'-•‘^^
m
A* ^•».^:.V 7 .

ti
ff:-

'•'•IV

€1 j1 iSfeaisqfqA

mi'-

M~ :II jdl>ad<|qA
$'

':. /.
.
.^^

or tin xrbil2>qc|A^

eslctet

i®PBB7r7;S™'? ,S -Tru'f -. v>:fc. _ ,

7 -

iB?
Infroduction

This document describes the procedures and instrumentation used by the National
Institute of Standards and Technology (NIST) aluminum ultrasonic reference
in calibrating

blocks. The Ultrasonics Group, in the Automated Production Technology Division of the
Manufacturing Engineering Laboratory, performs this calibration service for technological
and industrial customers. The calibration is based on the American Society of Testing and
Materials (ASTM) recognized practice aluminum reference blocks,
for calibrating
designated E127-92a; Standard Practice for Fabricating and Checking Aluminum Alloy
Ultrasonic Standard Reference Blocks [1]^ Instructions for requesting calibrations of
ultrasonic reference blocks are contained in Appendix I of this document.
The NIST ultrasonic reference block calibration service was originated in the late
1970's, and is NBS Technical Note 924 entitled "Procedures for the
described in the
Calibration of ASTM
E127-Type Ultrasonic Reference Blocks." [2]. At that time the
application of one master block for use as a reference standard was considered to be a
temporary or interim measure. Today NIST uses this master block to set the ultrasonic
system sensitivity in all calibrations of ultrasonic reference block sets. In addition, there
are several newly implemented practices employed for checking the calibration system
characteristics, both over a period of years and in preparation for a particular calibration.

This current document updates the NBS Technical Note 924 in describing these practices

for our calibration procedure.

^ Steel and Titanium Ultrasonic reference blocks are also calibrated by NIST. While E127-92a does not
apply to these types of blocks, the instrumentation and some of the procedures described in this document
are the same. ASTM E428: Fabrication and Control of Steel Reference Blocks used in Ultrasonic
Inspection covers these types of blocks.

1
E127-Tvpe Ultrasonic Reference Blocks

The reference artifacts described in E127 are 7075-T6 aluminum alloy, 50.8 mm (2
in)2 diameter cylindrical blocks. Following NIST policy, we give quantities in SI units
with in-lb units in parentheses. The ASTM El 27 standard generally gives the in-lb units
first. The blocks are of varying length and have flat-bottom-holes (FBH) drilled into their

bottom surfaces. A schematic of this type of block is shown in Fig. 1. In describing these
types of blocks the following labeling convention is used in El 27: the hole diameter,
expressed in 64ths of an inch, and the metal path length, expressed in lO-^ of an inch, are
separated by a dash. For example, a block with a 5/64 in diameter hole (also referred to as
a No. 5 hole) and a 0.50 in metal path would be designated as 5-0050. Blocks with this
labeling are commonly used; however, El 27 gives equivalent hole diameters and lengths in
SI units. This labeling convention is used throughout tWs document, but quantities will
generally be given in accord with NIST policy.

Measurement Facility

A thorough understanding of the ultrasonic calibration equipment is essential in

order to maintain a consistent and repeatable calibration service. Changes in the
calibration system can be detected, corrected, and even anticipated when one has a
sufficient understanding and historical charting of the instrumentation involved. The past
two decades have provided us with the data and experience necessary to understand
which measurement factors are important for reliably calibrating aluminum ultrasonic
reference blocks.
All reference block calibrations are done in a 300 L tank, as shown in Fig. 2.

Distilled water, with a nominal temperature range of 20-24 °C, is used as the coupling
fluid. Previous research has shown that temperature variations in this range do not affect
the calibration results when all components are in thermal equilibrium [3]. The water is

kept chlorinated in order to prevent algae growth, etc., and while the conductivity of the
water is increased as a result, the short-term exposure of the blocks to the water is not
harmful to the physical characteristics of the blocks. In addition, a filtration system is

periodically used to keep the coupling fluid clean of small debris such as dust. Calibration
blocks are placed in the tank on a set of large aluminum steps, machined to a high
which keep all of the blocks' fi-ont-surfaces at the same height. This minimizes
precision,
the amount of vertical transducer manipulation needed during the calibration procedure.
A large moveable stage is attached to the top of the tank, as shown in Fig. 2. The
ultrasonic transducer is connected to by means of a 98 cm long ultra-high-
this stage
fi’equency (UHF) coaxial metal rod. The transducer can be manually manipulated in two
perpendicular planes through the use of an angulator. The transducer also has degrees of

^English units are given first with SI units in parentheses by ASTM in E127 in describing reference
blocks and transducer characteristics. This is contrary to the U.S. national policy of using metric units
only. However, this document both the English and metric equivalent of measured quantities in a
lists

manner consistent with NIST policy to provide coherency for those who are familiar with E127. NIST is
currently working with ASTM to promote the use of metric units only.

2
freedom in the -X, -Y, and -Z directions, with movement in these directions controlled

manually or with motorized linear actuators contained within the stage. This setup is

sufldcient for properly positioning the transducer during all phases of the calibration.

Instrumentation

Figure 3 shows a schematic of the ultrasonic calibration instrumentation. Each

component of this setup is explained in detail below.^

Transducer
One of the most important components of the calibration system is the ultrasonic
transducer. Ultrasonic transducers convert electrical pulse-excitations into pulses of high-
frequency sound which are above the threshold of human hearing (in this case 5 MHz).
Pulse repetition rates of 200-1000 s*^ are typically used in this calibration. The ultrasonic
pulses are emitted from the transducer and travel through a water medium (i.e. coupling)
to the target of interest. At a material interface or discontinuity some of the ultrasonic
energy is back towards the transducer. The transducer then receives these
reflected
reflected echoes and converts them back into electrical signals for electronic detection.
For a given ultrasonic pulse incident on a target, the amplitude of the return echo depends
on the target size and shape, and on the acoustic properties of both the target and the
coupling medium. The measurement parameter of interest in the calibration of ultrasonic
reference blocks is the amplitude of the return echoes reflected from the flat-bottom-holes
drilled into the bottom of the blocks. In this particular calibration, the interface of interest
is a solid-air boundary between the solid aluminum and the FBH.
An Aerotech 5 MHz quartz immersion-type transducer possessing a 9.525 mm
(0.375 in) effective element size, historically designated "Lab-Standard-4" (LS-4) [4], has
been used in virtually all of NIST's ultrasonic reference block calibrations since the
inception of the calibration service. Quartz transducers, while not commonly used for
generic nondestructive testing due to their insensitivity, do have the advantage of
remaining very stable over long periods of use. In addition, quartz transducers are more
easily constructed to the same or similar performance characteristics, than are ceramic
units [4]. Similarity in performance between like-constructed transducers is important for
maint^ning widespread applicability of the ASTM E127-92a measurement standard when
using different equipment. The pulsed beam-profile characteristics of LS-4 have been
carefully and repeatedly measured over time using pulse-echo techniques with specific
targets. We note that LS-4 meets the requirements of section 11.5 of E127-92a in size,

operating frequency, and lateral and axial beam profiles. The transducer beam profiles,
taken on a 12.7 mm (0.5 in) diameter steel sphere target, are shown in Figs. 4-6. Figure

^Commercial equipment or materials are identified in this paper in order to specify adequately certain
calibration procedures. In no case does such identification imply recommendation or endorsement by the
National Institute of Standards and Technology, nor does it imply that the material or equipment
identified is necessarily the best available for the purpose.

3
4a shows lateral scans ofLS-4 taken at the transducer's near-field (Y^) length ^ The scans
2ire taken at four different rotational orientations of the transducer which are separated by
45° of rotation. Figure 4b shows a comparison of one of the 4a scans with one which was
taken by Chwirut in 1976 [2]. Figure 5 is a lateral scan of LS-4 taken at the transducer's
near-field minimum (1[") point. Figure 6 is an axial scan of LS-4 which clearly shows the
and points in the transducer's sound field. All of the beam profiles taken at the

present time are in good agreement with those obtained by Chwirut in the 1970's [2],
indicating that the transducer's response has not measurably changed with time.
To complement the pulse-echo scans, hydrophone scans have recently been made
of the transducer's beam profile. A thin polymer membrane, made of polyvinyldifluoride
(PVDF) in this case, can measure ultrasonic pulses traveling through it in either direction,
with minimal perturbation to the pulses themselves. Ultrasonic pulses measured by
traditional transducers by the transfer-function characteristics of the
are affected
transducer, resulting in a measured waveform that differs fi'om the actual waveform. In
contrast, the significantly greater bandwidth and closer impedance match to water of a
PVDF hydrophone allows for a high fidelity recording of the actual ultrasonic waveform.
A commercially available hydrophone possessing a circular element size of 0.3 mm^
(0.0005 in2) and a bandwidth > 20 MHz was used to map the ultrasonic beam intensity at
the transducer's near-field (I^) length. The transducer was moved slowly over a 645 mm^
(1 in^) area in a raster-scan fashion. The result, shown in Fig. 7, shows a good symmetric
beam profile which is fi'ee of significant irregularities.

Pulser/Receiver
The pulser/receiver currently used in NIST's calibration service is an Automation
Industries Inc. model M90 Reflectoscope. When this unit is used in conjunction with an
external LC tuning circuit, it provides pulse excitations of sufficient energy to drive the
requisite quartz transducers such as LS-4. The receiver has a coarse-adjustment
bandwidth extending to 10 MHz and a gain range of approximately 0-81 dB. This gain is

adjustableby a 20 dB-increment switch (range 0-40 dB), a 2 dB-increment knob (range 0-

30 dB), a 1 dB-increment switch (range 0-1 dB), and a continuous gain vernier (range 0-
10 dB.) This gain range is sufficient for adequately detecting the ultrasonic reflections
fi’om FBHs for the metal-paths and hole sizes specified in E127-92a. When making
appropriate gain changes during calibrations, the labeled gain values indicated on the M90
are not used. Rather, the amplitude of the reflected signal is used to measure gain
changes, as described later in this document.

transducer's near field length is the natural focus distance in an ultrasonic beam emanating from a
flat transducer. The sound field intensity monotonically decreases beyond this length, and is therefore
generally a better region for target placement. The near field length N of a circular spot source, in a
coupling fluid, can be approximated analytically by the expression:

4A
where D is the transducer's active element diameter and X is the wavelength of the ultrasound in the
coupling fluid. Note that X varies inversely with the frequency of the ultrasound.

4
 The receiver linearity of the M90 Reflectoscope is checked periodically, using the
setup shown in Fig. 8. A function generator is used to generate a signal pulse which has a
similar spectral content as a typical echo received by the M90 and LS-4 located 88.9 mm

(3.5 in, its 17 point), from the surface of a reference block target in water. This signal is

attenuated by an independently calibrated attenuation strip, and then detected by the M90
through its receiver port. The attenuation strip is calibrated using a regulated DG source
and has a standard uncertainty of ± 0.1 dB to ± 0.3 dB over the range of 1 dB to 60 dB.
A digital voltmeter measures the peak amplitude of the unattenuated input signal and the
subsequent attenuated signals. The detected peak amplitude is then compared to the value
computed using the unattenuated signal amplitude and the attenuation value. This is
repeated over a range of attenuation values, for each of several different receiver gain
settings on the M90. A typical linearity test for one receiver gain setting is shown in Fig.
9. The receiver portion of the M90 has a measured amplitude linearity of better than 1.5

% of full-scale (FS), over a range of amplitude values that exceed those normally used in
these calibrations. (While the M90 is linear within 1.5 FS up to 6 V, 0.3-5. 0 V has %
been defined to be the instrument's full scale linear range.)
In the calibration, the ultrasonic signal of interest is selected on the M90 screen for
measurement by positioning the M90's receiver gate. The peak amplitude of the signal
marked by this electronic gate is then detected by the M90 and delivered to a digital
voltmeter. An adjustable pulse-length knob on the M90 is kept at its maximum (full
clock-wise) position at all times. The M90 shock-excitation is then tuned with an external
tuning circuit described below.

PT-4 Tuning Circuit

The PT-4 tuning circuit indicated in the general calibration setup of Fig. 3 is used
as an external tuner for the shock-excitation pulses from the M90 Refleaoscope that drive
quartz transducers. The circuit is a commercial one, built by F. Baxter Co., and
distributed by Eastern NDT [5] for use with low capacitance transducers. A schematic for
this circuit is shown in Fig. 10. This circuit, which is external to the M90, is needed since

the M90 is a shock-excitation pulser without the internal tuning capability for the tuned
excitation specified in E127-92a. Together these two units very closely approximate the
performance (when used with the same quartz transducer) of the calibration
pulser/receiver unit used by NIST in the 1970's and 80's.

Experience has shown that an important aspect of a reliable ultrasonic calibration

system is maintaining the constancy of the frequency content of this excitation pulse. As a
matter of practice the excitation pulse of the calibration system is periodically checked,
and tuned if necessary. The experimental setup for this procedure is shown in Fig. 11.
The transducer is attached to the system in its normal way and immersed in water so as to
maintain the usual physical loading on the transducer. The M90 shock-excitation pulse is

observed on an oscilloscope and its frequency content is measured on a calibrated

spectrum analyzer. If necessary, the pulse length and tuning are adjusted, using the
control knobs (variable resistor for pulse length and variable capacitor for pulse tuning) in
the PT-4 circuit. The spectral content of this shock-excitation is tuned to have a center-
frequency of 5.00 MHz ±0.05 MHz. A typical excitation pulse from the M90, with the

5
PT-4 tuning circuit, and its corresponding frequency spectrum are shown in Fig. 12. In
the time domain this pulse resembles a damped one-and-a-half period sinusoidal pulse.
One variable which affects the frequency content of the excitation pulse reaching
the transducer is the length of the BNC cables and UHF rod between the pulser and
transducer. Longer cables result in a higher capacitance between the various pieces of
instrumentation used in the calibration. The higher capacitance affects the system's
of properly tuning the transducer's center frequency. As a result, great care has
capability
been taken to use the same cable lengths for all calibrations to ensure that the spectral
content of the excitation pulses remains constant. The cable lengths used in NIST
calibrations are indicated on the appropriate setups (Figs. 3 and 11.) These lengths, when
used in conjunction with the current calibration instrumentation, allow for a tuned center
frequency of 5.0 MHz.

RC Circuit/Digital Voltmeter
The positive peak amplitude of the detected ultrasonic pulse of interest from the
M90 detector unit is displayed on a digital voltmeter (DVM). This signal is filtered by an
RC circuit possessing a time constant of 0. 1 s, which reduces the effect of spurious noise
on the voltage readings by time averaging. This filtering improves the stability of the
signal reading on the DVM. Due to the long time constant of the circuit, any signal
maximizing (while aligning or normalizing the transducer for example) is usually done by
observing the signal on the M90 screen, and not by reading the DVM.

Calibration Procedures

Preliminary Settings
After the system's excitation pulse and linearity have been checked the calibration
can be performed. The following steps are taken for both Area-Amplitude (A-A) and
Distance-Amplitude (D-A) block calibrations. Appendbc n contains sample data-
acquisition formswhich show the detailed steps of each type of calibration.
is turned on at least one hour before data is taken, in order to
All instrumentation
allow for adequate instrument warm-up and stable voltage readings. Before being placed
in the tanks, all blocks are visually inspected, and any grease or oil removed from their
front surfaces. Nicks, scratches, rust, etc. which are on the front surfaces are noted for
each of the blocks, since these may affect the ultrasonic response of the block. This
includes any surfaces not appearing to meet the surface roughness criteria in section 10.2
of E127-92a [1]. Finally, the hole plugs on the bottom of each block are inspected to
make sure that they appear to be securely sealed and free of corrosion.
As shown in Fig. 2, blocks of varying metal-path-distance are placed in the water
on large aluminum steps. This keeps the blocks' front-surfaces at the same height, thereby
minimizing the need for adjusting the transducer-to-target distance during the calibration.
Once submerged the blocks' entry surfaces are lightly brushed to remove air bubbles which
might otherwise interfere with the ultrasonic beam and therefore affect the calibration
measurement. Back-lighting through the tank's transparent walls provides a means of
easily detecting these air bubbles.

6
 A 88.9 mm (3.5 in) long spacer-block is then placed on top of one of the blocks to
be calibrated. The transducer is lowered to the top of this spacer-block, which is

subsequently removed. The transducer is angulated for normalization to the front-surface

of the calibration block and is considered to be normal to the front-surface when a
maximum-amplitude front-surface-reflection is detected. The time position of the
reflected pulse is then marked on the display of the M90 Reflectoscope. This reference
point is used when repositioning the transducer in order to keep the correct and constant
water path for all blocks.

Distance-Amplitude (D-A) Sets

The system sensitivity is set using a master 5-0050 aluminum reference block,
designated Ultrasonic Reference Block- 1 (URB-1). After beam normalization to the
block's front surface, the transducer is moved laterally, in a transverse plane, until a

maximum-amplitude reflection from this block's FBH is detected. The system gain is then
adjusted until the reflection amplitude is 80% of full-scale. In this procedure, we note that
a master reference block is used instead of the appropriately sized steel balls which are
called for in E127-92a.^ After setting the initial sensitivity, a gain change which depends
on the hole size of the blocks may be needed. Gain settings for the respective hole sizes
are determined by their relative FBH areas. Thus, for a number 3 size hole D-A set the
gain would be increased by a factor of (25/9), while for a number 8 size hole D-A set the
gain would be decreased by a factor of (25/64) compared with a No. 5 hole set. Previous
research at NIST has shown that these procedural differences yield more precise and
meaningful reference block calibration data, in consonance with E127-92a [6]. In
addition, the gain is increased by (31 to 808.2 %
FS) halfway through a D-A set, in
dB
order to obtain an adequate scope display and remain in the system linearity range for the
echo responses from the longer metal-path blocks. All gain changes are made in the
following manner, where a 3 1 % to 80 % FS gain increase is used as an example: The
transducer is manipulated laterally until the peak amplitude of any appropriate reflection,
usually that from a FBH, decreases to 31 %
FS. The gain controls on the M90 are then
adjusted until the peak amplitude of this signal reads 80 FS. The peak amplitude during %
this gain adjustment is read on the DVM. Throughout this document, including Appendix
n, gain changes are expressed in terms of dB. These are simply unit conversions from the
actual % FS changes used in the gain adjustments. The actual labeled gain values
indicated on the M90 are not used for these gain adjustments.
A check-standard (CS) block is measured for each gmn setting, to verify that the
appropriate gain changes have been made
These check-standard blocks are also
correctly.
used as a means for determining the overall uncertainty of the measurement system, as
discussed below. The physical characteristics of the CS blocks are given in Table 1.

^The appropriately sized steel balls called for in ASTM-E127-92a are measured after setting system-
sensitivityon URB-1. These values can then be used by the customer as a reference value. Care must be
taken when using the ratio of these values as a "NIST-to-E127 conversion factor," since the different
target geometries have different scattering characteristics. These include differences in the spectral
content of the reflected pulses and thus differences in system linearity for the distinct targets. In addition,
detailed experiments at NIST show that better accuracy is obtained with URB-1 than with steel balls [7].

7
 During the calibration, the maximum echo-amplitude from each block's FBH is
read and recorded. The entire measurement scheme is done three times for each block in
the set. These three independent measurements yield an average and a check on
procedural repeatability for each block. A large variation for a particular block may be an
indicationthat one trial measurement was done incorrectly, and that additional
measurements are necessary. The block-response value reported to the customer is the
average value of these three trials.

Area-Amplitude (A-A) Sets

For Area- Amplitude sets, the system-sensitivity is set on the largest-hole block
(typically 8-0300) in the set. The echo-amplitude from the FBH of this block is set to
100% of the full-scale linearity of the system. Each of the smaller-hole blocks are then
read in turn, making gain changes of 8.2 dB after reading the No. 2 and 3 blocks in order
to keep the signal in the system's linearity range. As for D-A sets, tlu-ee independent
measurements are made of each block in the set. After all eight A-A blocks (a complete
set) have been measured, the 3-0300, 5-0300, and 8-0300 blocks are measured again,
using the D-A methodology, as shown in the A-A data-sheet in Appendix n. This is done
to determine whether the absolute response of these blocks is correct, as stated in section
lL8ofE127-92a[l].

Calibration Report Contents

After the measurements are completed, NIST provides a written report to the
customer, summarizing the calibration results. A sample calibration report is included in
Appendk m of this document. The typical calibration report will include the reduced
ultrasonic data in tabular and graphical form, a statement of block acceptance/rejection
with respect to the criteria set forth in E127-92a, and a statement regarding the NIST
calibration system standard uncertainties.

NIST Ultrasonic Calibration Databases

The NIST Ultrasonics Group has maintained a database of all calibrations

performed since the mid-1970's. In addition, CS data since that time has also been
recorded as a means of assessing the measurement uncertainty associated with this
calibration.

Calibration Database
Figure 13 shows a comparison of the NIST calibration database, which includes all

calibrated aluminum reference block values, with the minimum and maximum limits
prescribed by E127-92a, for metal paths from 12.7 mm (0.5 in) to 152.4 mm (6.0 in) and
for No. 3, 5, and 8 hole sizes. The average value for each metal-path in the database is
indicated (•), along with their ± two standard deviation (2a) ranges for all blocks with this
metal-path, to indicate their 95% confidence intervals. The minimum and maximum limits
specified in E127
from the NIST database
are also indicated (
fall
— ). For all but the 0575 blocks, the average values
within the specifications called for by E127-92a. For the
larger metal paths there is a trend of the NIST average values to lie closer to the minimum

8
E127-92a values. Table III lists the average and standard deviation of all aluminum
reference blocks calibrated by NIST, for each metal path. The number of blocks
calibrated for each hole size is also indicated. It is worth noting that the large number of
44.5 mm (1.75 in) and 57.2 mm (2.25 in) metal path blocks is artificially large, since these
blocks are measured twice in each D-A calibration.
The ASTM committee E-07-06, which is responsible for maintaining the E127-92a
document, is currently examining the possibility of using the empirical NIST database for
establishing new E127-92a recommended values. If this is to be done a fit of the NIST
valuesmay be used. Here we fit the average NIST values to an empirical relationship
which was previously found by Chwirut to give the best fit for this type of data-set [7].
That empirical relationship is of the following form:

y(x)= (1)
(a + bxy
where x is the material path length and y(x) is the echo-amplitude fi'om the FBH. This
function's form has a sound theoretical justification, since the e"®" term takes into account
the material attenuation a, and the (a + bx)^ term is the predicted relation for the
reflection amplitude fi’om a disc residing in the far-field of the sound beam [7]. The
coefficient a is very small for 5 MHz ultrasound in aluminum, while the a and b
coefficients depend on the transducer characteristics, the water pathlength, transmission
and reflection coefficients, sound velocities of the coupling fluid and material, etc. Using
our database and a least-squares optimization algorithm [8] we have calculated numerical
values for these coefficients. These are presented in Table 11. These values describe the
NIST aluminum block calibration data, using LS-4 at a water path of 88.9 mm (3.5 in).

Figure 14 shows the NIST average values for each metal-path, and a plot of equation (1)
using the calculated values of the three coefficients. The calculated attenuation coefficient
a= 0.047 dB/cm (0.0138 NpAm) is compared with the value a= 0.02 dB/cm for 5 MHz,
as presented by McMaster for AL17ST (0.13 mm grain size) [9].
Check-Standard Database
Figures 15-20 show plots of the check-standard values measured fi’om 1976 to
present. The raw data (o) and ± 2a ( — ) values are indicated. Note that each point is the
average of the three measurements taken for each calibration. The ± 2a values represent
the expanded measurement uncertainties of the calibration system, with a representing the
combined standard uncertainty [10]. The measurement uncertainty of a measured block
value then depends on the particular block's metal-path distance and hole size, as indicated
in Table 1, since different check-standards are used with different-sized blocks.

Conclusion

Here, we have summarized the NIST calibration of ASTM El 27-type ultrasonic

aluminum reference blocks. The equipment used in the calibration, and the methods used
for maintaining the equipment characteristics have been described. We have also briefly
outlined the calibration procedures themselves, and have given an overview of the NIST

9
calibration database which contains all No. 3, 5, and 8 hole-size aluminum blocks
calibrated by NIST since 1976.

Acknowledgments

The authors would like to thank Dr. David Gilsirm for his help with the numerical
least-squares algorithm. We would also like to acknowledge the many NIST and ASTM
E-07-06 readers of the draft copy of this document, including Dr. Donald Eitzen, Mr. Bill
Crupe, Mr. Howard Van Valkenburg (New Fairfield, CT.), and Mr. Jerry Green (Alcoa
Aluminum Co., Alcoa Center, PA.). All of their comments and suggestions for this
document are greatly appreciated.

10
 References

[1] Standard Practice for Fabricating and Checking Aluminum Alloy Ultrasonic
Standard Reference Blocks, E127-92a, in Annual Book of ASTM Standards Philadelphia, .

Pennsylvania: American Society for Testing and Materials; 1992.

[2] Chwirut, D.J., Sushinsky, G.F., and Eitzen, D.G., Procedures for the Calibration of
ASTM E 127-Tvpe Ultrasonic Reference Blocks . Nat Bur. Stand. (U.S.) Tech. Note
924; (September 1976).

[3] UnpubHshed NIST Data.

[4] Chwirut, D.J., and Boswell, G.D., The Evaluation of Search Units used for
Ultrasonic Reference Block Calibrations . Nat Bur. Stand. (U.S.) Internal Report NBSIR
78-1454; (February 1978).

[5] Eastern NDT Inc., 200 N. Main Street, Suite 212, Hopewell, Virginia 23860

[6] Sushinsky, G.F., et al. . Improved Ultrasonic Standard Reference Blocks . NBSIR
76-984; (November 1976).

[7] Chwirut, D.J., Recent Improvements to the ASTM-tvpe Ultrasonic Reference

Block System . NBSIR 79-1742; (April 1979).

[8] Atkinson, K.E., An Introduction to Numerical Analysis . John Wiley and Sons, NY
(1978).

[9] McMaster, R.C., Nondestmctive Testing Handbook Roland Press Company,

. NY
(1959).

[10] Taylor, B.N., and Kuyatt, C.E., Guidelines for Evaluating and Expressing the
Uncertainty of NIST Measurement Results. NIST Technical Note 1297 (January 1993).

11
THIS PAGE INTENTIONALLY LEFT BLANK

12
Appendix I: Procedures for Requesting Ultrasonic Reference Block

Calibrations

Interested parties should call the technical contact listed below, before sending

block sets to be calibrated. The technical contact will inform the customer of the cost of

the calibration, and any other pertinent information. The customer should then send a

company purchase order and the block set(s) to the address below. The purchase order

should preferably be sent prior to the block set(s) and should clearly state the type of

calibration to be performed, the cost, purchase order number, the bill-to office, and the

name, address, and phone number of the person responsible for the block set(s). No
calibrations can be performed until NIST has received and processed this purchase order.

Therefore, the tum-around time for ultrasonic reference block calibrations can be

shortened if NIST receives the purchase order before the block sets. Normally, customers

can expect a four week tum-around time, although this time can be reduced in special

circumstances. A table listing the various reference block calibration services offered by

the NIST Ultrasonics Group is presented on the next page.

Technical Contact:
Mr. John A. Slotwinski
Building 233 / Room AMT
N.I.S.T.
Gaithersburg, Maryland 20899-0001
(301)975-2171

13
 NIST Ultrasonics Group Reference Block Calibration Services

Test No. DescriDtion of Services Fee ($)

26030S Special Test of Area Amplitude Aluminum 830

Reference Blocks - Set of Eight Blocks
26040S Special Test of Area Amplitude Titanium or At Cost
Steel Ultrasonic Reference Blocks - Set of
Eight Blocks
26050S Special Test of Distance Amplitude 1130
Aluminum Reference Blocks - Set of
Fifteen Blocks
26060S Special Test of Distance Amplitude At Cost
Titanium or Steel Ultrasonic Reference
Blocks - Set of Fifteen Blocks
26070S Special Test of Distance/Area Amplitude At Cost
(Basic Set) Aluminum, Titanium or Steel
Ultrasonic Reference Blocks - Set of Eight
Blocks

These prices are subject to change. Consult the latest edition of the NIST Calibration

Services Users Guide and Fee Schedule (NIST Special Publication 250, January 1993), or

contact the technical contact listed on the previous page for more information.

14
Appendix II: Data-AcquisitioP Fonns used in NIST Calibrations of
ASTM E127-tvpe Reference Blocks

COMPANY: AREA-AMPLITUDE CAUB. NO.

S/N: TECHNICIAN
DATE
BLOCK Trial 1 Trial 2 Trial 3
8-0300 100% 100% 100%
7-0300
6-0300
5^)300
4-0300
3-0300
INCREASE GAIN 8.2 dB
3-0300
2-0200
INCREASE GAIN 8.2 dB
2-0300
1-0300
URB-1 80% 80% 80%
INCREASE GAIN 8.6 dB
INCREASE GAIN 8.2 dB
3.175 mm (1/8 in) BALL
CS-2
3-0300
URB-1 80% 80% 80%
INCREASE GAIN 8.2 dB
7.938 mm (5/16 in) BALL
CS-4
50300
URB-1 31.5% 31.5% 31.5%
INCREASE GAIN 8.2 dB
17.463 mm (11/16 in) BAa
8-0300

15
COMPANY: DISTANCE-AMPLITUDE CAUB. NO.
S/N: ALUMINUM NO. 3 HOLE TECHNICIAN
DATE
BLOCK Trial 1 Trial 2 Trial 3
URB-1 80% 80% 80%
INCREASE GAIN 8.9 dB
3.175 mm (1/8 in) BALL
CS-1
3-0050
3-0063
3-0075
3-0088
3-0100
3-0125
3-0175
3-0225
INCREASE GAIN 8.2 dB
CS-2
30175
3-0225
30275
3-0325
3-0375
3-0425
3-0475
30525
30575

16
COMPANY: DISTANCE-AMPUTUDE CAUB. NO.:
S/N: ALUMINUM NO. 5 HOLE TECHNICIAN:
DATE:
BLOCK Trial 1 Trial 2 Trial 3
URB-1 80% 80% 80%
7.938 mm (5/16 in) BALL
(DS-S
5-0050
5-0063
50075
5-0088
50100
50125
50175
50225
INCREASE GAIN 8.2 dB
CS-4
50175
50225
50275
5-0325
50375
50425
50475
&0525
50575

17
COMPANY: DISTANCE-AMPUTUDE CAUB. NO.
S/N: ALUMINUM NO. 8 HOLE TECHNICIAN
DATE
BLOCK Trial 1 2
Trial Trial 3
URB-1 31.3% 31.3% 31.3%
17.463 mm (11/16 in) BAU
CS-5
8-0050
8-0063
8-0075
8-0088
8-0100
8-0125
Br0]75
8-0225
INCREASE GAIN 8.2 dB
CS^
8-0175
8-0225
8-0275
8-0325
8-0375
8-0425
8-0475
8-0525
8-0575

18
Appendix HI: Sample Distance-Amplitude Calibration Report

TN-OOOOOO-93
J.A. Slotwinski/bam
September 23, 1993

U.S. DEPARTMENT OF COMMERCE

NATIONAL mSHTUTE OF STANDARDS AND TECHNOLOGY
GAITHERSBURG, MD 20899

REPORT OF CALIBRATION

OF

One Set of Aluminum No. 5 D-A Blocks

Block Manufacturer S.N. 0001

Requested by

Any Company
P.O. Box 0001
Anywhere, USA

The set of fifteen aluminum reference blocks was calibrated according to the procedures
of the NIST document entitled "NIST Calibration of ASTM E127-Type Ultrasonic
Reference Blocks." In these procedures, the ultrasonic system sensitivity is based on the

echo amplitude from the National Institute of Standards and Technology (NIST) master
flat-bottom-hole (FBH) block of size 5-0050, rather than the steel reference balls called
for in ASTM E127-92a. Gain settings for the respective hole sizes are in turn determined
by FBH areas. Research at NIST has indicated that these procedural
their relative
differences yield more and meaningful reference block calibration data in
precise
consonance with ASTM E127-92a. (Using these procedures, the echo amplitude response
value of the 7.9 mm (5/16 in) steel ball was 70.1 FS.) %
The calibration results are given in Fig. For each reference block, the percent of full
1.

scale (% FS) average of three readings is graphed and tabulated. Note that the system
gain was increased from 31 to 80% FS, after the first measurement of the 57.2 mm (2.25
in) metal-path block. After taking into account NIST procedural differences and the
combined standard uncertainty of the system, the echo amplitude responses of all blocks
except 5-0575 fell within the range of the E127-92a established values. The 5-0575 block
was slightly lower than the established values of E127-92a.

19
We also report NIST system combined standard uncertainties using a NIST master set of
aluminum FBH blocks as check standards. These (total) system combined standard
uncertainty values are determined for each receiver gain setting that is used for each range
of blocks. They represent random and some systematic errors associated with the NIST
system. (We note that other systematic errors or changes may produce differences greater
than the uncertainties reported here.) Based on check standard data accumulated since
NIST mm mm
1976, the

(0.50 to 2.25 in)] is

f 2 2\—
system expanded uncertainty for the short blocks [12.7 to 57.2

of the block response readings, and for the long blocks [44.5
V64.4y
^ 2,4
mm to 152.4 mm (1.75 to 6.00 in)] after a gain increase is of the block response

readings. (The stated expanded uncertainty is twice the combined uncertainty (2a value),
corresponding to a 95% confidence interval.)

We note that the values reported herein apply to the blocks calibrated only with the
ultrasonic system and instrument settings used at NIST. The National Institute of
Standards and Technology cannot assure that the customer's blocks will have these same
values when measured in the customer's facility on another day. The dependence of the
echo amplitude on the particular ultrasonic system used is well documented, and is

available from NIST upon request.

20
 I 1
m
D
CM

Amplitude
S
(%FS) M 00 ^ lO te <0 U CM o to 04 CO CO
(5 O i 8 oo.*»
48.8
40.6

S CO
S85
35.4

a •o
CM 8 CD
Echo
s
«o

I1 11
i
u> 0 U) c > «o to
8 !5 to
E
fl
§8 c
r>.

5 o o O i o c5S 8 8 S S 8 8
BlocklO

3 li
1
C
1z 5

CALIBRATION

BLOCK

V) Number

iP >* o
C
E o
D a
C E
E o ID
D o
< o
REFERENCE o Block
OQ

o o

o o

ULTRASONIC

O
CO
O Q
"O
Q
io
O
^ O
CO
O
CM
O

(Sd 7o) ©ptnildoiv 0M03

21
 Table I: Dimensional and ultrasonic characteristics of NIST check-standard blocks.

Block Block Size Average Echo 2a Applicable to

Designation Amplitude (% FS)
(% FS)
CS-1 3-0225 22.9 2.4 short No. 3 blocks
CS-2 3-0175 79.6 4.8 long No. 3 blocks
CS-3 5-0075 64.4 2.2 short No. 5 blocks
CS-4 5-0375 29.4 2.4 long No. 5 blocks
CS-5 8-0175 29.3 1.2 short No. 8 blocks
CS-6 8-0325 35.6 1.4 long No. 8 blocks

Table 11; Calculated values of numerical coefficients in Eq. (1), using NIST database.

Coefficient Computed Numerical Computed Numerical

Value (metric) Value (in-lb)

a 0.047 dB/cm 0.0138 (Np/in)

a 0.0836 (-) 0.0836 (-)

b 0.023 (cm-i) 0.0584 (in-i)

22
Table HI: Average, standard deviation, and number of blocks calibrated for each metal
path in the NIST database, since 1976.

Block ID Number Average (% FS) a (% FS) Number of Blocks

0050 75.3 3.6 116
0063 70.3 3.5 73
0075 61.9 3.1 115
0088 55.9 2.6 74
0100 49.6 2.3 93
0125 40.4 1.5 75
0150 31.9 2.3 34
0175 27.8 1.5 151
0200 23.5 2.3 29
0225 20.2 1.3 150
0250 17.7 2.1 24
0275 15.5 1.1 95
0300 14.4 2.4 103
0325 12.3 1.0 76
0375 10.0 0.9 79
0400 9.7 1.2 14
0425 8.2 0.9 74
0475 7.0 0.9 74
0500 7.7 1.4 14
0525 6.2 1.0 74
0575 5.5 0.9 73
0600 7.4 2.9 31

23
 Ultrasonic Wave Entry Surface

Flat-Bottom-Hole

Figure 1: Schematic of generic ASTM E127-92a-type aluminum reference block which

shows the position and orientation of the flat-bottom-hole (FBH) relative to the ultrasonic
wave entay surface.

24
Figure 2: Photograph of ultrasonic calibration system, immersion tank, and block holder.

25
Figure 3: Ultrasonic Instrumentation Setup for calibration of ASTM E127-92a type
aluminum reference blocks, with cable lengths indicated.

26
 90

(a)

0 Degrees

” " " 45 Degrees

— * ” 90 Degrees
135 Degrees

(b)

1993

• 1976

Lateral Position (mm)

Figure 4: (a) Lateral scans of LS-4 taken using a 12.7 mm (0.5 in) steel sphere target at
89.9 mm (I^ point) water path distance. The scans are taken at four different rotational
orientations of the transducer, which are separated by 45° of rotation, (b) Comparison of
one of the 4a scans with one which was taken by Chwirut [2] in 1976. The good
agreement between the 1993 continuous scan data and the 1976 discrete data is one
indication of LS'4's stability over time.

27
 o
o
oo
o s s o o
CO
o
CM
O O
(SJ 7o) spn«|duiv 0403

28
 target,

sphere

steel

in)

(0.5

mm scan.

12.7
the

a on

using
indicated

taken

are

points
LS-4

of
Yf
scan

and

Axial

6:

transducer's

Figure

O
O Q
^ Q
lO Q
^ O
CO
O
CN
O O

(Sd %) opniiidoiv oqoa

29
 10
map
its by
at This

separated
field

mV.

sound 40
are
irregularities.

to

LS-4 contours

the

of dark
corresponding
significant

in^)]
of
(1 heavy

contour free

The
mm^
is
645
dark
which

hydrophone.

(central) profile

[approximately

first
PVDF beam

a the
map
by with symmetric

Intensity measured

good

7: increments,

as a
Figure
point,
shows

mV

[(dais/ui two) dajs/uiui uonisoj -

30
Figure 8: Arrangement used to measure the M90‘s receiver amplitude linearity.

31
 JD
2
13

o o
0
•D

8 peak

ideal

8 and

measured

8
the

4>
8 D
3 showing

id •i—

a
E
amplitude.

8
< measurement

o peak

<D
CL
T3
8 O
predicted

linearity

O
T3 of
0)
o. M90

8 typical
function

a
A as

9:
8
amplitudes

Figure

8.00 88888888
l<'dlO‘^COCN»—
(A) epruiiduiv ^loed pejnsDe|/\|
o
8

32
"Output" Jack "Receive" Jack

"Transmit" Jack

Figure 10: PT-4 tuning circuit schematic. This circuit is used to tune the M90 shock-
excitation.

33
Figure 11: Setup for checking the finequency
content of the M90 shock-excitation.

34
 (a)

Time (fis)

Frequency (MHz)

Figure 12: (a) Typical shock excitation pulse (tuned) and its (b) corresponding frequency
spectrum on a linear vertical scale.

35
 )

two

NIST {± and
2a
by ±
minimum

the

calibrated

the

represent

blocks
represent

all
(X)

the

contains
(
and
The

value,
which

database.

amplitude
database

the

in
block echo

length E127.92a.

average
path
reference

ASTM

the metal

in
ultrasonic

represent
each

specified

for

NIST (•)
as

The
The deviation)

limits

13: 1976.

maximum
standard

Figure
since

o o o o o o O o
0*001
g g R g g g o
CO
C)
CM
c5 o

(Sd %) epninduJV 0403

36
 best-fit

II.
corresponding

Table

the
from

and

(o)
coefficients

values

calculated
average

the
database

and

1
NIST
Eq.

the using

of
plot made

A is
14:
which

),
Figure —

8 S O
CO
O S S o o
CO
O
(N

<Sd 7o) epntnduiv oqoa

37
 O)
1.

Cd
Check-Standard

o
o>
indicated.

from

00
00 1
are

Inc. CO measured

Tabulation
(D limits

154-4
CO
00
> ) —

Industries,
CO amplitude (
TO
s/n c
00 (D deviation

CO echo

Data 3-0225,

CM
o the

Automation
00 0 of
standard

E
o
00
i- charting

±
two

CS-1 and

(©)
00 Historical

h- points

15:

CO data

Figure

The

o o
o o)
o
(30
o
r^
o
in
o o

(Sd %) 0pnj!|dijuv 0 L|03

38
 2.

CsJ Check-Standard

o>

O indicated.

from

abulation Inc.
CO are

00 measured

limits

154-4 )
CD
—

T
Industries,
00 amplitude (
s/n

deviation

00 echo
3-0175,

Data
Automation the
CVJ standard

00 of

o
00
charting

±
two

CS-2
and

Historical
(o)
00
points

16:

CO data

Figure

The

ooooooooooo
O0)00N-CDLOM-CQC\J'i-
(Sd %) 0 pni!idijuv OM03
39
 3.

a>

eg Check-Standard

O)

o
o>
indicated.

from

are

abulation 00 measured

CO limits
Year)
122
)
Inc. CO
(
T s/n
00 amplitude

(Calendar deviation

Trienco, 00 echo

5-0075,
Data
the

CM of
standard

00
Time
two
charting

±
CS-3 and

Historical
(o)

points

17:

data

CO
Figure

The

(Sd %) epnindoiv 0403

40
 4.

O)

CM Check-Standard

o>

o from
indicated.

are

abulation
00 measured

00 limits

122 ) —

CD
Inc.
00 (
T s/n
amplitude

deviation

Trienco,
00 echo

5-0375,
Data the

standard

CM of
00
two

o
charting

±
CS-4 00 and

Historical
(o)

00 points
r^
18:

data

CD
Figure

The

ooooooooooo
OG)CX)r^CDLO^COCM^
(Sd %) 0pni!|dijuv 0^03
41
 O)

OJ
CD

O
O)

abulatlon
00
00

123
Inc.
CO
T s/n
00

Trienco,
00
Data 8-0175

CM
00

CS-5

CO

"M"

(Sd %) Qpnindujv 0403

 6.

O)

CVJ Check-Standard

O)

o
O)
indicated.

from

are

00 measured

00 limits
Tabulation

123 ) —

Inc. CO
00 co amplitude (
s/n "D
c
o deviation

00 CO
echo

Trienco,
Data
8-0325 the

CM
00
0 of
standard

e
o H charting

±
two

CS-6 00 and

Historical
(o)

00
points
r--
20:

data
co
Figure

The

ooooooooooo
oocDh-com^cocvj^
(Sd %) apniiidoiv 0403
43